Although most lung cancer patients are treated with a planning target volume to cover the motion margin, it is desired to reduce the motion margin for a more conformal treatment. Technological improvements in the software and hardware have allowed us to deliver synchronized radiation or to track a known trajectory of tumor motion. However, the exact position of the tumor is difficult to detect in real time. One or a few fiducial markers, trackable by optical cameras are not reliable and reproducible surrogates of the internal tumor motion; x-ray fluorescent imaging delivers radiation dose to a large patient volume and is not able to detect small lung tumors or tumors with low radiographic contrast. In this proposed project, a more robust and mechanistic correlation between the patient surface and the lung tumor is proposed based on a combination of 4DCT, 4DMRI and surface optical tracking. 3D imaging modalities will be utilized to create boundary conditions for mechanical modeling and validation of the model. To transfer the proposed technology into treatment room, optical tracking will be developed to provide surface conditions instead of MRI.